Upgrade your account to a Community Supporter account and remove most of the site ads.

Reply to thread

Message: <blockquote data-quote="FrogReaver" data-source="post: 9104762" data-attributes="member: 6795602">IMO.  No issues so far.My first major point is that specificity matters!The premises:Person A has agency at point x1, x2, x3, x4, ... , xnPerson B has agency at point x1, x2, x4, ... , xnPlayer Z has agency at point x1, x2, y1, y2, y3, ... , ypPlayer F has agency at point x1Player G has agency at points x1, x2The first question is, why didn't player B have a choice at point x3?  Possibility - the choice player B made at point x2 allowed him to bypass point x3.  Would this result in him having less agency than player A or the same?  There's certainly less points where he has agency, but a greater percentage of them.  Which brings me to my next major point.The issue isn't simply that there's no method to measure agency in a meaningful way, but there's also the question of which metric to use for agency in the first place - of which we have at least 2 (and possibly more).Let's also look at player B and player Z.  This is where player B's choice at x2 and player Z's choice at x2 completely diverge so that player B and player Z no longer share points.  Assuming we can map all this out (the measurement problem), we could come up with a particular answer for Player B and Player Z and after picking which metric to use for agency we could answer that question.Now consider Players F and Player G.  Suppose that player F and G both have so few points where they have choice because after their respective points end, they did reach other points but had no choice at them.  We could measure both of these and conclude that player 1 had more agency under the percentage method and player 2 had more agency under the additive method.  But I think intuitively either of those options just feels wrong.  Neither of these players actually get to affect the game itself as after their first move or 2 they are essentially on a railroad.  Which is where I want to establish my next major point.Agency over a point in a game (and agency over all points in a game) and agency in respect to the game itself aren't the same things.  Or more generally - agency always must be defined in respect to something.Side Bar: Another interesting question is what counts as a particular point.  For example can point z1 be equal to points x1 and x2 combined?  If so couldn't one for example have agency with respect to z1 and x2 but not x1?  Or when we are talking particular points is what is actually being suggested the existence of atomic points that cannot be further broken down?My final major Point.I'd suggest there's a 3rd and better metric for agency.  Agency is always about whether one has a choice with respect to something.  Why do I propose it's a better metric?  Because it better aligns with our intuitions.  It doesn't require us to determine what's an 'atomic point'.  It doesn't require us to pick percentage or total.  Instead we can determine the answer by simply asking the following question: did player X have the capacity to make a choice that could affect change with respect to thing Y.  Or as I've previously shorthanded it - Agency is Binary.I'm not going to go immediately into this after my wall of text above, but I think there's a good argument that more choices with respect to a particular X doesn't of itself constitute more agency.I'd suggest it's up to us to choose the metric by which to measure agency, and if we choose the 'agency is binary' metric we can measure it.  I've laid out a case for why I think it's the better choice, but ultimately it's a choice we have to make.</blockquote>

Verification